Skip to content

Commit 45bb245

Browse files
V4belopsiff
authored andcommitted
FROMLIST: net: skbuff: propagate shared-frag marker through frag-transfer helpers
maillist inclusion category: bugfix CVE: CVE-2026-46300 Two frag-transfer helpers (__pskb_copy_fclone() and skb_shift()) fail to propagate the SKBFL_SHARED_FRAG bit in skb_shinfo()->flags when moving frags from source to destination. __pskb_copy_fclone() defers the rest of the shinfo metadata to skb_copy_header() after copying frag descriptors, but that helper only carries over gso_{size,segs, type} and never touches skb_shinfo()->flags; skb_shift() moves frag descriptors directly and leaves flags untouched. As a result, the destination skb keeps a reference to the same externally-owned or page-cache-backed pages while reporting skb_has_shared_frag() as false. The mismatch is harmful in any in-place writer that uses skb_has_shared_frag() to decide whether shared pages must be detoured through skb_cow_data(). ESP input is one such writer (esp4.c, esp6.c), and a single nft 'dup to <local>' rule -- or any other nf_dup_ipv4() / xt_TEE caller -- is enough to land a pskb_copy()'d skb in esp_input() with the marker stripped, letting an unprivileged user write into the page cache of a root-owned read-only file via authencesn-ESN stray writes. Set SKBFL_SHARED_FRAG on the destination whenever frag descriptors were actually moved from the source. skb_copy() and skb_copy_expand() share skb_copy_header() too but linearize all paged data into freshly allocated head storage and emerge with nr_frags == 0, so skb_has_shared_frag() returns false on its own; they need no change. The same omission exists in skb_gro_receive() and skb_gro_receive_list(). The former moves the incoming skb's frag descriptors into the accumulator's last sub-skb via two paths (a direct frag-move loop and the head_frag + memcpy path); the latter chains the incoming skb whole onto p's frag_list. Downstream skb_segment() reads only skb_shinfo(p)->flags, and skb_segment_list() reuses each sub-skb's shinfo as the nskb -- both p and lp must carry the marker. The same omission also exists in tcp_clone_payload(), which builds an MTU probe skb by moving frag descriptors from skbs on sk_write_queue into a freshly allocated nskb. The helper falls into the same family and warrants the same fix for consistency; no TCP TX-side in-place writer is currently known to reach a user page through this gap, but a future consumer depending on the marker would regress silently. Fixes: cef401d ("net: fix possible wrong checksum generation") Fixes: f4c50a4034e6 ("xfrm: esp: avoid in-place decrypt on shared skb frags") Suggested-by: Sabrina Dubroca <sd@queasysnail.net> Suggested-by: Sultan Alsawaf <sultan@kerneltoast.com> Suggested-by: Ben Hutchings <ben@decadent.org.uk> Cc: stable@vger.kernel.org Signed-off-by: Hyunwoo Kim <imv4bel@gmail.com> Link: https://github.com/v12-security/pocs/tree/main/fragnesia Link: https://x.com/v12sec/status/2054491454064746629 Link: https://lists.openwall.net/netdev/2026/05/13/79 Link: https://lore.kernel.org/all/20260513041635.1289541-1-vakzz@zellic.io/ Link: https://lore.kernel.org/all/agRfuVOeMI5pbHhY@v4bel/ Signed-off-by: Wentao Guan <guanwentao@uniontech.com>
1 parent 642e1d5 commit 45bb245

3 files changed

Lines changed: 8 additions & 0 deletions

File tree

net/core/gro.c

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -214,10 +214,12 @@ int skb_gro_receive(struct sk_buff *p, struct sk_buff *skb)
214214
p->data_len += len;
215215
p->truesize += delta_truesize;
216216
p->len += len;
217+
skb_shinfo(p)->flags |= skbinfo->flags & SKBFL_SHARED_FRAG;
217218
if (lp != p) {
218219
lp->data_len += len;
219220
lp->truesize += delta_truesize;
220221
lp->len += len;
222+
skb_shinfo(lp)->flags |= skbinfo->flags & SKBFL_SHARED_FRAG;
221223
}
222224
NAPI_GRO_CB(skb)->same_flow = 1;
223225
return 0;
@@ -245,6 +247,8 @@ int skb_gro_receive_list(struct sk_buff *p, struct sk_buff *skb)
245247
p->truesize += skb->truesize;
246248
p->len += skb->len;
247249

250+
skb_shinfo(p)->flags |= skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG;
251+
248252
NAPI_GRO_CB(skb)->same_flow = 1;
249253

250254
return 0;

net/core/skbuff.c

Lines changed: 3 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2214,6 +2214,7 @@ struct sk_buff *__pskb_copy_fclone(struct sk_buff *skb, int headroom,
22142214
skb_frag_ref(skb, i);
22152215
}
22162216
skb_shinfo(n)->nr_frags = i;
2217+
skb_shinfo(n)->flags |= skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG;
22172218
}
22182219

22192220
if (skb_has_frag_list(skb)) {
@@ -4289,6 +4290,8 @@ int skb_shift(struct sk_buff *tgt, struct sk_buff *skb, int shiftlen)
42894290
tgt->ip_summed = CHECKSUM_PARTIAL;
42904291
skb->ip_summed = CHECKSUM_PARTIAL;
42914292

4293+
skb_shinfo(tgt)->flags |= skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG;
4294+
42924295
skb_len_add(skb, -shiftlen);
42934296
skb_len_add(tgt, shiftlen);
42944297

net/ipv4/tcp_output.c

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -2389,6 +2389,7 @@ static int tcp_clone_payload(struct sock *sk, struct sk_buff *to,
23892389
todo = min_t(int, skb_frag_size(fragfrom),
23902390
probe_size - len);
23912391
len += todo;
2392+
skb_shinfo(to)->flags |= skb_shinfo(skb)->flags & SKBFL_SHARED_FRAG;
23922393
if (lastfrag &&
23932394
skb_frag_page(fragfrom) == skb_frag_page(lastfrag) &&
23942395
skb_frag_off(fragfrom) == skb_frag_off(lastfrag) +

0 commit comments

Comments
 (0)